Primarystructures ofchickenconevisual pigments: rhodopsins

Proc. Nati. Acad. Sci. USA
Vol. 89, pp. 5932-5936, July 1992
Biochemistry

Primary structures of chicken cone visual pigments: Vertebrate
rhodopsins have evolved out of cone visual pigments

(vislon/cDNA doning/molecular evolution)

TOSHIYUKI OKANO*, DAISUKE KoJIMA*, YOSHITAKA FUKADA*, YOSHINORI SHICHIDA*,
AND T6RU YOSHIZAWAt

*Department of Biophysics, Faculty of Science, Kyoto University, Sakyo-ku, Kyoto 606-01, Japan; and tDepartment of Applied Physics and Chemistry,
The University of Electro-Communications, 1-5-1, Chofugaoka, Chofu, Tokyo 182, Japan

Communicated by John E. Dowling, March 20, 1992

ABSTRACT The chicken retina contains rhodopsin (a rod retinas, might belong to one of the known groups or to a
visual pigment) and four kinds of cone visual pigments. The previously undisclosed one.
primar structures of chicken red (iodopsin) and rhodopsin
have been determined previously. Here we report isolation of In addition, chicken green-sensitive cone pigment (chicken
three cDNA clones encoding additional pigments from a green) combines characteristics normally found only in
chicken retinal cDNA library. Based on the partial amino acid rhodopsins with those in cone pigments. Chicken green (508
sequences of the purified chicken visual pigments together with nm; ref. 9) has an absorption maximum closer to the
their biochemical and spectral properties, we have identified rhodopsins (around 500 nm) than does any other cone pig-
these clones as encoding the chicken green, blue, and violet ment that had been sequenced. Furthermore, it, like
visual pigments. Chicken violet was very similar to human blue rhodopsin, has a high affinity for concanavalin A (8, 9),
not only in absorption maximum (chicken violet, 415 nm; suggesting that its oligosaccharide chains are similar to those
human blue, 419 nm) but also in amino acid sequence (80.6% of rhodopsin. On the other hand, unlike rhodopsin, chicken
identical). Interestingly, chicken green was more similar (71- green and the other chicken cone pigments are positively
75.1%) than any other known cone pigment (42.0-53.7%) to charged at neutral pH as shown by the lack of affinity for
vertebrate rhodopsins. The fourth additional cone pigment, DEAE-Sepharose (8, 9). In addition, chicken green and other
chicken blue, had relatively low similarity (39.3-54.6%) in cone pigments are unstable in the presence of hydroxylamine
amino acid sequence to those of the other vertebrate visual even in the dark (8, 9). Thus, chicken green possesses
pigments. A phylogenetic tree of vertebrate visual pigments properties of both rod and cone pigments. These unique
constructed on the basis of amino acid identity indicated that characteristics prompted us to determine the primary struc-
an ancestral visual pigment evolved first into four groups ture of chicken green and the other chicken cone pigmentsi
(groups L, S., Ml, and M2), each of which includes one of the
chicken cone pigments, and that group Rh including vertebrate Although our previous work (10) showed that chicken red
rhodopsins diverged from group M2 later. Thus, it is suggested should be classified into group L because of its similarity to
that the gene for scotopic vision (rhodopsin) has evolved out of human red and green, the present study clearly shows that
that for photopic vision (cone pigments). The divergence of chicken blue and green should be classified into two different
rhodopsin from cone pigments was accompanied by an increase groups. That is, an ancestral cone-like pigment evolved first
in negative net charge of the pigment. into four groups, from one of which rhodopsins diverged
later. This implies that scotopic vision appeared after animals
Cone photoreceptor cells of higher vertebrates act under acquired color vision.
daylight conditions, and their characteristic color sensitivi-
ties are attributed to the visual pigments having unique MATERIALS AND METHODS
absorption spectra. Humans have three kinds of cone visual
pigments (1) with absorption maxima at 558 nm (human red), Library and Probes. A chick retinal cDNA library was
531 nm (human green), and 419 nm (human blue) (2), which constructed in Agtl1 phage vector by using 5 pug of poly(A)+
have been classified into two groups on the basis of amino RNA from 1-day-old chick retinas and a cDNA cloning kit
acid similarity (3): long-middle wavelength-sensitive pigment (Amersham).
(group L; human red and green) and short wavelength-
sensitive pigment (group S; human blue). A recent investi- In the primary screening, a 72-nucleotide probe termed
gation of the primary structure of cone visual pigments in RV72 (5'-GATCATCACCACCACCATGCGGGA-
monkeys supports this idea (4). CACCTCCTTCTCCGCCTTCTGCGTCGACTCCGAC-
TCCTTCTGCTGAGC-3', which is antisense for Ala-Gln-
Microspectrophotometric experiments (5-7) showed that, Gln-Lys-Glu-Ser-Glu-Ser-Thr-Gln-Lys-Ala-Glu-Lys-Glu-
unlike primates, some vertebrates have a tetrachromatic Val-Ser-Arg-Met-Val-Val-Val-Met-Ile, the loop V-VI region
color vision. In fact, four kinds of cone visual pigments were of chicken red; ref. 10) was used. To characterize the cDNA
extracted from chicken retinas (8, 9). We determined absorp- clones isolated in the primary screening, we used three
tion maxima of the pigments, which are located at 571 nm additional probes: LYS30 (5'-AGGATTGTAGATGGCA-
(chicken red), 508 nm (chicken green), 455 nm (chicken blue), GAGCTCTTGGCAAA-3', which is antisense for Phe-Ala-
and 415 nm (chicken violet) (9). It is interesting to examine Lys-Ser-Ser-Ala-Ile-Tyr-Asn-Pro, the helix VII region of
whether the fourth cone visual pigment, lacking in primate chicken rhodopsin; ref. 11), CT30 (5'-AGCAGAGCT-
GTCCTCATCGCCCAGCGGGTT-3', which is antisense for
The publication costs of this article were defrayed in part by page charge Asn-Pro-Leu-Gly-Asp-Glu-Asp-Thr-Ser-Ala, the carboxyl-
payment. This article must therefore be hereby marked "advertisement" terminal region of chicken rhodopsin; ref. 11), and BL35
in accordance with 18 U.S.C. §1734 solely to indicate this fact. [5'-GTCATGGCCTT(G or C)CC(A or G)CA(G or C)AC-
CATCTTCATGATGCA-3', which is antisense for Cys-Ile-

MThe sequences reported in this paper have been deposited in the

GenBank data base (accession nos. M92037-M92039).

5932

Biochemistry: Okano et al. Proc. Natl. Acad. Sci. USA 89 (1992) 5933

Met-Lys-Met-Val-Cys-Gly-Lys-Ala-Met-Thr, the carboxyl- where n(i) denotes the number ofpositively chargeable amino
acids arginine (R), lysine (K), and histidine (H); and n(j)
terminal region of human blue; ref. 3]. denotes the number of negatively chargeable amino acids
The synthetic probes were radiolabeled with [y-32P]ATP by aspartic acid (D), glutamic acid (E), and tyrosine (Y). Since
cysteine residues are often modified posttranslationally (16,
the use of Megalabel kit (Takara Shuzo, Kyoto). Inserts of 17), the dissociation of sulfhydryl groups was not considered.
The isoelectric point (pl) of a pigment was estimated by
the isolated cDNA clones were also used as probes after solving the equation, NC(pI) = 0. The following pKa values
for the amino acid side chains in water at 250C were used
being labeled with [a-32P]dCTP by the use of a random primer (3.65, 4.25, 6.00, 10.08, 10.53, and 12.48 for D, E, H, Y, K,
and R, respectively; ref. 18).
labeling kit (Takara Shuzo).
RESULTS AND DISCUSSION
Cloning and Sequencing. Plaques of Agtl1 phage were
Isolation of cDNA Clones. The DNA sequence encoding the
transferred to nylon membranes (Hybond-N+; Amersham). loop V-VI region of chicken red was selected as a screening
probe (RV72) because many vertebrate visual pigments have
A hybridization with the synthetic oligonucleotide probes similar sequences in this region. Then, 2.0 x 105independent
oligo(dT)-primed cDNA clones were screened with the use of
was carried out in a hybridization buffer composed of 5 x SSC radiolabeled RV72, resulting in the isolation of 46 positive
clones. From the clones obtained, those encoding chicken
(1 x SSC = 0.3 M sodium chloride/30 mM sodium citrate, pH red or rhodopsin were eliminated by using plaque hybridiza-
7.0), 10% formamide, 10 mM sodium phosphate, 0.25% tion. The clones encoding chicken red or rhodopsin were
sterilized skim milk (Difco), 10%o dextran sulfate, and 100 pug identified by strong hybridization only with probe ANECO
of denatured salmon sperm DNA per ml at 420C. In the case (10) or with both CT30 and LYS30, respectively. Then, we
obtained a clone AY2E [761 base pairs (bp)] hybridizing with
of the hybridization with labeled cDNA inserts, the concen- both probes BL35 and LYS30 but not with CT30 or ANECO.
The deduced amino acid sequence of AY2E (217 amino acids)
tration of formamide in the hybridization buffer was raised to
40%o. After the hybridization, the nylon membranes were was very similar to that of human blue. The upstream
washed with 2x SSC containing 0.1% SDS at 500C. Phages
fragment of AY2E (5' end to the Pst I site; 145 bp) was used
comprising DNA hybridized with the probe were identified as a probe for further screening. Briefly, 1.4 x 106 indepen-
dent oligo(dT)-primed cDNA clones and 4 x 105 independent
by autoradiography or a Bio-Image analyzer (FUJIX BA100, random-primed cDNA clones were screened to isolate more
than 20 positive clones from which 2 clones were isolated:
Fuji Film). ARc2 (1408 bp, a full-length clone of AY2E encoding chicken
violet; see below) and AF1(1711 bp, a clone encoding chicken
The cDNA inserts excised from the phage DNAs were blue) hybridizing weakly with AY2E. Next, 1.4 x 106 inde-
pendent oligo(dT)-primed cDNA clones were again screened
subcloned into the pBluescript II KS(+) (Stratagene) plasmid by using AF1 as a probe, resulting in isolation of AF7G (2451
bp, a clone encoding chicken green) and AF9Rh (1326 bp, a
vector. The subclones were subjected to a nested deletion full-length clone of chicken rhodopsin, whose sequence had
already been reported; ref. 11).
(12). Both strands of all the subclones were sequenced by the
dideoxy chain-termination method (13) with the aid of Se- Characteristics of the Deduced Amino Acid Sequences. The
deduced amino acid sequences of the three clones, ARc2,
quenase version 2.0 in the presence of single-stranded DNA AF1, and AF7G (Fig. 1), had several similarities and differ-
ences (Table 1). First, highly conserved amino acids were
binding protein (United States Biochemical). found in loop III-IV and V-VI regions, which in bovine
rhodopsin are known to interact with transducin (26). Thus,
Peptide Analyses of Chicken Green and Blue. Chicken green these cone pigments would also probably interact with trans-
was purified as described (9). Chicken blue was purified by ducin, as was shown for chicken red (27). Second, the amino
acid sequences of these three proteins carry a net positive
subjecting the chicken blue-enriched fraction (9) to car- charge at neutral pH (see below). This result is consistent
boxymethyl-Sepharose column chromatography (unpub- with the chicken cone pigments lacking an affinity for DEAE-
Sepharose at pH 6.6 (8, 9). Thus, chicken cone pigments are
lished data). Highly purified chicken green or blue (0.4 mg in contrast with rhodopsin, which is negatively charged as
each) was digested by lysyl-endopeptidase/Achromobacter shown by adsorption to the DEAE-Sepharose (8, 9).
lyticus protease I (8 ,ug each; Wako Biochemicals, Osaka) in
10 mM Tris-HCl, pH 9.0/4 M urea/2 mM dithiothreitol for 24 RNA Blot-Hybridization (Northern Blot) Analysis. A North-
hr at 37°C. After the digest was centrifuged (90,000 x g for ern blot analysis of chick retinal poly(A)+ RNA (4 ,g per
lane) was performed by using the isolated cDNA inserts as
20 min) to remove insoluble materials, the supernatant was hybridization probes (Fig. 2). AF7G and AF1 hybridized with
2.8-kilobase (kb) and 3.1-kb RNA, respectively. ARc2 hy-
applied to a Cosmosil SC18 P-300 column (4.6 x 150 mm; bridized with a single faint band (3.7 kb). AF9Rh showed a
Nacalai Tesque, Kyoto) equipped with a HPLC system main band (1.6 kb) and a faint band (2.5 kb), suggesting
(model 600E; Waters). Proteolytic fragments were eluted multiple transcription sites in chicken retina as have been
with a linear gradient (5-90%) of acetonitrile in 0.05% tri- observed in rat, human, and frog opsin genes (28).
fluoroacetic acid at a flow rate of 1 ml/min. Amino acid
sequences of the purified peptides thus obtained were deter- Assignment of the Clones to the Pigments. The partial amino
mined by a gas-phase automated sequencer (model 477A;
Applied Biosystems) and a phenylthiohydantoin amino acid acid sequence (Thr-Glu-Val-Ser-Ser-Val-Ser-Ser-Ser-Gln-
analyzer (model 120A; Applied Biosystems).
Val-Ser-Pro-Ala) of a carboxyl-terminal fragment obtained by
The Phylogenetic Tree. A phylogenetic tree of visual pig- digestion of the highly purified chicken green with A. lyticus
ments based on the ratio of amino acid substitution was protease I perfectly coincided with the amino acid sequence
constructed according to a neighbor-joining method (14). The
amino acid identities (% identity) between each pair of the
sequences were calculated at every position excluding posi-
tions where gaps exist in either of the compared sequences.
The evolutionary distance [k = -ln(1 - K)] was estimated by
calculating the amino acid difference [K = 1 -(% identity)/
100]. For estimating the probabilities (P) of occurrence of the
tree topologies, the bootstrap resampling procedure (15) was
repeated 500 times.

Calculations of a Net Charge and an Isoelectric Point. The
net electric charge (NC) of the opsin moiety at given pH
[NC(pH)] was estimated by the following formula:

R KH 1

NC(pH) = n(i) 1 + 1oPH-PKaY)

D Y n(j) 1
10pKa(j>pH'
1 +
+

5934 Biochemistry: Okano et al. Proc. Natl. Acad. Sci. USA 89 (1992)

PosiioninBovineRhodopsin 1 10 20 30 40 60 70 80 90 100
II
I.I

Chicken Red MAAWEAAFAARRRHE-EEDTTRDS&VFTYTNSN )NTRGPFEGPNYHIAPRWVYNLTSLWMIFVVAASVFTNGLVLVATWKFKKLRHPLNWILVNLAVADLGETVIASTISVINQISGYF
..QQQ.SLQRL.G.HPQDSY. .S.QS.......
Human ...QQ.SLQRL.G.HPQDSY....-I S................... H. ....T...............V .................. I.V.............

Human S. _I................................. ..................... A ..AV..VY...

Fish .GDQ.GD.V ...... GD .... E.. KD ATC. F.. V. TV SA. I.. .L. .LL C. .FF...

Fish G101 H_.PV.....HN EA.V ...N .. I. .A. .. L. I...
Fish G103 .. L. F. .VF...
YNHAD.PV ..... .E.A.V A D...... A ..I.. I.. S ..I .A .I.

Chicken Green MNGTEGINFYVPMS NKNKTGVVRSPFEYPQYYLAEPWKYRLVCCYIFFLISTGLP INLLTLLVTFKHKKLRQPLNYILVNLAVADLFMACFGFTVTFYTAWNGYF
Rh
Chi,dTnt~ernoy Rh DF....... ....... .... .LA ...Y. ... SALAA.M ... LV.F.V.F F. .VQ..T .. L.. .M.N ... VL. M. .SM....

Q- .LL.F.V.F.YIQ...... -------Q LT .SM..... .....V VFG. ..T.M.
....................... FSALAA.M.M.

Bovine Rh F......P.. .... QFSMLAA.M.L. ..F.........A ......... .Y..V. ..
.ML.F L. VFG T.L. .SLH...

Human Rh ..... P... .... PF. VFL.A ...QFSMLAA.M.L. .VL.F. Y..VQ. T. L. VLG. ..S.L. .SLH...

Chicken Blue MHPPRPTTDLPEDF_YIPMALDAPNITAL-SPFLVPQTHLGSPGLFRAMAAFMFLLIALGVP INTLTIFCTARFRKLRSHLNYILVNLALANLLVILVGSTTACYSFSQMYF

Chicken Violet MSSDDDFYLFT NGSVPGPWGP YHIAPLPWAFYLQTAFMGIVFAVGTPLLNAVVLWVTVRYKRLRQPLNYI LVNISASGFVSCVLSVFVVFVASARGYF
.I.SV_ V..VA.L..KV.
Human Blue MRK .._EEE.. ...K A....T..LI.F.... ..V.FG. .LL.IF ... P CN...

110 120 130 140 150 160 170 180 190 200 220 230

III IV VI

Chicken Red ILGHPMCWEGYTVSACGITALWSLAI ISWERWFVVCKPFGNIKFDGKLAVAGILFSWLWSCAWTAPP IFGWSRYWPHGLKTSCGPDVFSGSSDPGVQSYMWLMVTCCFFPLAI I ILCYLQVWLAIR
G VR. I.AVL
Human Red V.. L ..L .A IV .A. ... Y I II. M.
Human
Fish Green V.. L LG .M VR. .A IV..A ..I.AAV Y I ..........S V.
Fsh
Fish R007 F. .F..AT AG.. ..TV. V V M.T .V.T.V ..AV.C E I..I .I.... IG

G1OI IF. .W V TV V. W.AG. .I V.AII.CT E A... IT. .L IL. .S ... ..IF. .S. .H

G103 V.. IF. V. T V V W.AG..I.A.T.AII.CT E ..IIL. .SV .IF. .N. .H

Chicken Green VFGPVGCAVEGFFATLGGQVALWSLWLAIERYIWCKPMGNFRFSATHAMMGIAFTWVMAFSCAAPP LFGWSRYMPEGMQCSCGPDYYTHNPDYHNESYVLYMFVI HF I IPVWVIFFSYGRLICKVR
Lagmprey Rh .... TM.SI .E . .I. GN . .I .V. . .. .LA. V .I. .. L. .NFN . .V. . V. .LV.F.I ..C.C .. L.T.K
Chicken Rh ... VT..YI. EI. V..V.V.. S .GEN ..I.V. .S.I. MA. . I. . I.... LK.EIN F.I....VI.M. .LA... C. .N.V.T.K
Bovine Rh .T..NL .El.V.. S. .. GEN. .I.V.. LA . V .I. .. I. ...PHEETN F.I .V.. LI .... C. .Q.VFT.K
Human Rh
...T...NL .E.V..S... .GEN. .I.V.. LA . A .I.. .L.. .... LK.EVN . .F.I .I. V. .T .MIIC ...C. .Q.VFT.K

Chicken Blue ALGPTACKIEGFAATLGGMVSLWSLAWAFERFLVI CKPLGNFTFRGSHAVLGCVATWVLGFVASAPP LFGWSRY I PEGLQCSCGPDWYTTDNKWHNESYVLFLFTFCFGVP LAI IVFSYGRLLITLR

Chicken Violet VFGKRVCELEAFVGTHGGLVTGHSLAFLAFERYI V ICKPFGNFRFS SRHALLWVVATWL IGVGVGLPP FFGWSRYMPEGLQC SCGPDWYTVGTKYRSEYYTWFLF IFCF IVP LSL I IFSY SQLLSALR
Human Blue ... RH..A..G.L ..VA. K .T. .L. . .. ...SI. FI ..S.
C T... .R. .K

240 250 ~ VI 270 280 290 ViI 310 320 330 340 348

Chicken Red AVAAQQKESESTQKAEKEVSRMVVVMIVAYCFCWGPYTFFACFAAANPGYAFHPLAAALPAYFAKSATIYNP I IYVFMNRQFRNC_ LQLFGKKVDDGSEVSTS. RTEVSSVSNSSVSPA (362)
Human Red K. ...T FF..V..M V..._ L.SASK .. (365)
.

Human Green K. ...T VL.F ..A . P ... .M.F. V. . L.SASK .. (365)
Fish R007 T. .Q.
. M. . . V . V...M .. (341)
Fish G101 Q.Q... D..L.FIV .... AS. .T.S.V....MWM........... M.......E.A....G.-T .G..TA.
Fish G103 Q.Q... D .L.FIL .... AS. .T.S.L . ......W.... ... ..E.A. ...G. (353)
T ... TA. (355)

Chicken Green EAAAQQQESATTQKAEKEVTRMVILMVLGFMLAWTPYAVVAFWIFTNKGADFTATLMAVPAFFSKSSSLYNP I IYVLMNKQFRNCMITTICCGKNPFGDEDVSSTVSQSKTEVSSVSSSQVSPA (355)
Launorey Rh A. S.V.... .ALVC.V.. ... Y ... HQ.S .. .F.TL ... A.. .A ... V.-I..L..... L ..DESGA.T ...... T.P.... (353)
Chicken Rh .I............I....IA..L.I.C..V.....S.. ...Y ... Q.S. .GPIF.TIA .. A...AI ... .IV.L.... ..L ... T.-AG...T. T... (351)
Bovine Rh .................I.A..LI.C..L... ...I...Y. ..HQ.S. .GPIF.TI .. A.T.AV ...V. .IM. V.LL...... L..DEA.T._V.... T. .. A.. (348)
Human Rh .I. .IA.LIC.V. .S ... ....HQ.SN.GPIF.TI. ...A. .AAI...V..IM .L ......... L ..DEA._ATV.... T. .. A.. (348)

Chicken Blue AVARQQEQSATTQKADREVTKMVVVMVLGF LVCWAP YTAFALWVVTHRGRSFEVGLASIPSVF SKSSTVYNPVIYVLMNKQFRSCMLKLLFCGRSPFGDDEDVSGSSQATQVSSVSSSHVAPA (361)

Chicken Violet AVAAQQQESATTQKAEREVSRMVVVMVGSFCLCYVPYAALAMYMVNNRDHGLDLRLVTIPAFFSKSACVYNP I IYCFMNKQFRACIMETVCG}CPLTDDSDASTSAQRTEVSSVSSSQVGPT (347)
Human Blue ............................... ................................. .Q. .KM._ ._AM .E. ._TC.S._K ... T...T... N (348)

FIG. 1. The alignment of the primary structures of cone and rod visual pigments. The amino acid residues are expressed by the one-letter
amino acid code. Deduced amino acid sequences ofchicken red (10), human red (3), human green (3), fish putative cone pigments (19, 20), chicken
green, lamprey rhodopsin (21), chicken rhodopsin (11), bovine rhodopsin (22, 23), human rhodopsin (24), chicken blue, chicken violet, and human
blue (3) are aligned visually to optimize similarity. For simplicity, the sequence positions refer to those of bovine rhodopsin. Breaks in the
sequence indicated by lines denote deletions. Dots in human red, human green, and fish pigments indicate the identical amino acids to chicken
red. Dots in lamprey, chicken, bovine, and human rhodopsins indicate the identical amino acids to chicken green. Dots in human blue indicate

the identical amino acids to chicken violet. Putative transmembrane domains are shown by the lines above the alignment.

deduced from AF7G. Accordingly, the protein (355 amino cells but also those single cone cells with a deep-yellow oil
acids) encoded by AF7G was assigned to be chicken green. droplet (29) in which chicken green is present (30, 31). (ii)
Chicken green, which was very similar (73.2% identical) to Chicken green and rhodopsin showed a higher affinity for
chicken rhodopsin, displayed two potentially palmitoylated concanavalin A-Sepharose than did chicken red, blue, and
cysteines in the carboxyl-terminal region and two potentially violet (8, 9), suggesting chicken green shares with rhodopsin
glycosylated asparagines in the amino-terminal region (Table similar amino acid-bound carbohydrates.
1). However, the other two clones (AFi and ARc2) had only
On the basis of the partial amino acid sequence of the
one cysteine and one asparagine residue. These characteristics
of the deduced sequence of chicken green are consistent with seventh transmembrane region of chicken blue (Ser-Ser-Thr-
the following biochemical results. (i) An antiserum raised Val-Tyr-Asn-Pro-Val-Ile-Tyr-Val-Leu-Met-Asn-Lys), which
against bovine rhodopsin reacted with not only chicken rod was determined as described above, the second clone AFi
was assigned to encode chicken blue (absorption maximum,

Table 1. Characteristics of the clones and the deduced amino acid sequences of chicken visual pigments

Chicken visual pigment

Characteristics Red Green Rhodopsin Blue Violet

Clone ANECO AF7G AF9Rh AF1 ARc2
1.6 2.8 1.6, 2.5 3.1 3.7
Size of mRNA, kb
571 508 503 455 415
Pigment 362* 355 361 347
40,338 39,%5 351t 39,655 38,718
Absorption maximum, nm 9.6 9.4 9.9 9.5
100,100 62,900 39,325 52,100 66,000
Deduced amino acids, no. N31 N2,N15 5.9 N24 N12
K309 K296 K303 K291
Calculated molecular mass, Da C123-C200 C110-C187 51,700 C117-C194 C105-C182
Calculated isoelectric point None C322, C323 N2,N15 C330 C317
Calculated E at 280 nmt
K296
Potential glycosylation site(s)§ C110-C187
Chromophore-binding residue§ C322, C323
Intramolecular disulfide bond§
Possible palmitoylated residue(s)§

*From literature (10).
tFrom literature (11).

tCalculated by using the molar extinction coefficients (E) of tryptophan (5559 M-1 cm-1) and tyrosine (1197 M-1 cm-1) (25).
§Presumed by observations concerned to bovine rhodopsin (single-letter amino acid code is used).

Biochemistry: Okano et al. Proc. Natl. Acad. Sci. USA 89 (1992) 5935

Pigment Red Green Rh Blue Violet M

Clone Neco FRG FQRh F1 Pc2

4m' - 9.49kb a)
- 7.46kb
a - 4.40kb 00L)

- 2.37kb z

- 1.35kb

*- 0.24kb pH

FIG. 2. Northern blot analysis of chicken retinal poly(A)+ RNA FIG. 4. The net charge of the pigments. The net charge was
with the cDNA clones of chicken visual pigments. Blots of glyoxy- calculated as a function of pH for every pigment molecule (see
Materials and Methods).
lated poly(A)+ RNA (4 gg per lane) from chick retinas were hybrid-
rod and cone pigments. Thus, in higher vertebrates, rod and
ized with the radiolabeled insert DNAs (1.0 x 106 dpm) in the cone pigments are negatively and positively charged, respec-
hybridization buffer containing 47% formamide, washed with 2X tively, at neutral pH. Although the physiological significance
SSC containing 0.1% SDS at 50TC. Lane M shows molecular size of the molecular charge is not clear yet, some critical differ-
markers (0.24- to 9.5-kb RNA ladder) from BRL. ences in the photoresponses between rod and cone cells
(pigments) might be attributed to the opposite charges of the
455 nm; ref. 9). The primary structure of chicken blue (361
amino acids) showed relatively low similarity (39.3-54.6%) to pigment molecules.
any other known vertebrate visual pigment (Fig. 3). Molecular Evolution of Visual Pigments. The amino acid

As we have not yet succeeded in purifying chicken violet, sequences of chicken visual pigments were compared with
no direct sequence data are available at present. However, other vertebrate pigments (Fig. 1), and the percent identities
the assignment that the third clone (ARc2) encodes chicken in amino acid sequence were calculated (Fig. 3). Both figures
violet is strongly supported by the following two facts. (i) clearly show that the known vertebrate visual pigments can
ARc2 mRNA was less abundant than AF1 mRNA, as shown be classified into four groups (long, short, and two kinds of
by the Northern blot analysis (Fig. 2). This is consistent with middle-wavelength pigment groups: L, S, and M1/M2, re-
there being less (=20%; ref. 9) chicken violet in the retina spectively).
than chicken blue. (ii) The amino acid sequence deduced
from ARc2 was very similar (80.6% identical, Fig. 3) to that A phylogenetic tree of the vertebrate visual pigments (Fig.
of human blue, whose absorption maximum (419 nm; ref. 2) 5) was constructed by the neighbor-joining method (14) based
is close to that of chicken violet (415 nm; ref. 9). upon amino acid identity (Fig. 3). The reliability of the tree
topology was evaluated by bootstrap resampling (15). All of
Net Charge of the Pigments. It should be noted that many the tree topologies generated by resamplings 500 times were
of the positively charged residues in chicken green (Lys-36, the same as that shown in Fig. 5 except for two regions. (i)
Arg-38, Lys-64, Arg-225, Lys-229, and Lys-279) were re- Probability of the occurrence of the relationship among pig-
placed with noncharged residues as seen in bovine rhodopsin ments in group L (chicken red, human red and green, and three
(Fig. 1). Furthermore, noncharged residues in chicken green fish pigments) illustrated in Fig. 5 is relatively low (P = 0.81).
(Gln-122, Ala-150, and Pro-196) were replaced with glutamate (ii) Group M1 might be joined directly with group S (P = 0.30)
instead of group M2 (P = 0.70) as illustrated in Fig. 5.
residues as seen in bovine rhodopsin. These replacements
brought a striking shift of the isoelectric point from basic (pI The root of the vertebrate pigments (node A in Fig. 5) was
= 9.4) to acidic (pI = 5.9) (Table 1). When the net charges of determined by a tree including invertebrate rhodopsins (not
the cone and rod visual pigments were plotted against envi- shown). The bootstrap probabilities of occurrences of the
ronmental pH (Fig. 4), the cone pigments clearly split from tree topology with the deepest root of vertebrate pigments
rhodopsins except for lamprey rhodopsin, which lay between were 0.91 for the branch between nodes B and E (the present

g^iqS Ml grOuPIM t_ goqL

Human Chen Ctdon Human Chickn LaWey Clicim Fish Ash Fish Humm Humean cticin
L.P Blue Aied Bks Rh Rh Rh Green G103 G101 R0-07 G-r-ee-n Re-d R- -id 0
chcken Red 41.9 42.3 41.8 42.0 43.1
43.1 44.6 r 7&2 80.4 82.0 83.3 85.0 100

Human Red 42.1 42.3 42.5 44.7 45.3 45.0 45.9 7±0 73.9 77.7 95.9

Human Green 43.3 43.8 42.5 45.9 46.2 46.2 47.4 73.2 747 76.2 100

Fsh R007 40.1 40.4 42.6 41.7 40.7 40.4 42.3 77.7 80.2 100

Ash G101 41.0 39.6 40.5 40.8 40.8 40.5 42.0 94.3 100

Ash G103 41.6 40.5 39.3 40.2 40.8 40.5 42.6 10

ChienGren 47.8 49.4 53.7 71.0 73.2 75.1 100 FIG. 3. The percentage of amino acid
identity among several typical pigments.
Lamprey Rh 47.5 48.3 54.6 81.0 82.1 The values were calculated at every po-

Chiden Rh 46.6 46.2 51.1 873 1 sition excluding those positions where

Human Rh 46.2 46.6 51.9 100 gaps exist in either of the two sequences.
Values greater than 70o are in boldface
hickeinBske 49.4 49.9 1100
characters.
ChinMvbm 80.6 100

IHuman Blue lt

5936 Biochemistry: Okano et al. Proc. Natl. Acad. Sci. USA 89 (1992)

Fish R007

D.019 FhGo

Fish G

Da

L0.g0w0f Lamprey Rhodopsin FIG. 5. A phylogenetic tree of visual
pigments constructed on the basis of the
_____p Chicken Rhodopsin amino acid identity. The tree is shown
with branch lengths calculated from the
0.072 Human Rhodopsin
evolutionary distances (k). The deepest
root of the tree (node A) and the branch

length between nodes A and B were
determined by a tree including inverte-
brate rhodopsins (not shown). For sim-
plicity, some of the mammalian
rhodopsins (bovine, sheep, and mouse)
whose amino acid sequences are very
similar to human rhodopsin were
omitted.

topology), 0.08 for a branch between nodes B and F, 0.01 for coordination fund of the Science and Technology Agency of the
a branch between nodes C and D, and zero for the other Japanese government, and by a grant from the Human Frontier
Science Program.
branches.
1. Wald, G. (1967) Nature (London) 219, 800-807.
Chicken red (10), human red (3), human green (3), and 2. Dartnall, H. J. A., Bowmaker, J. K. & Mollon, J. D. (1983) Proc. R.
three kinds of putative fish visual pigments (19, 20) are
Soc. London Ser. B 220, 115-130.
clustered in group L, while chicken violet and human blue (3) 3. Nathans, J., Thomas, D. & Hogness, D. S. (1986) Science 232, 193-202.
are in group S. In contrast, neither chicken blue nor green can 4. Neitz, M., Neitz, J. & Jacobs, G. H. (1991) Science 252, 971-974.
be clustered with other known cone pigments. Chicken blue 5. Harosi, F. I. & Hashimoto, Y. (1983) Science 222, 1021-1023.
should be classified in a previously unreported group termed 6. Bowmaker, J. K. & Kunz, Y. W. (1987) Vision Res. 27, 2101-2108.
"group M1." Most interestingly, chicken green belongs to a 7. Hawryshyn, C. W. & Harosi, F. I. (1991) Vision Res. 31, 567-576.
large family (group M2) in which many kinds of vertebrate
rhodopsins (group Rh) are included (see below). The phylo- 8. Yen, L. & Fager, R. S. (1984) Vision Res. 24, 1555-1562.
genetic tree demonstrates that an ancestral visual pigment 9. Okano, T., Fukada, Y., Artamonov, I. D. & Yoshizawa, T. (1989)
evolved first into four groups (groups L, S, Ml1, and M2), each
of which includes one of the chicken cone pigments. Biochemistry 28, 8848-8856.
10. Kuwata, O., Imamoto, Y., Okano, T., Kokame, K., Kojima, D., Mat-
All of the sequenced cone visual pigments can be placed in
either group L or group S with the exception of chicken blue sumoto, H., Morodome, A., Fukada, Y., Shichida, Y., Yasuda, K.,
and chicken green, whose sequences are reported herein. Shimura, Y. & Yoshizawa, T. (1990) FEBS Lett. 272, 128-132.
Which diverged earlier, rod and cone pigments or groups L 11. Takao, M., Yasui, A. & Tokunaga, F. (1988) Vision Res. 28, 471-480.
and S, has been a moot question (32). The present determi- 12. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Cloning: A
nation of the primary structure of chicken green, which is Laboratory Manual (Cold Spring Harbor Lab., Cold Spring Harbor, NY),
closely related to that of vertebrate rhodopsin, should give a 2nd Ed.
definite point of divergence between rod and cone pigments 13. Sangar, F., Nicklen, S. & Coulson, A. (1977) Proc. Natl. Acad. Sci. USA
(node D in Fig. 5). Thus, it is evident that vertebrate 74, 5463-5467.
rhodopsins (group Rh) diverged from group M2 after the 14. Saitou, N. & Nei, M. (1987) Mol. Biol. Evol. 4, 406-425.
divergences of the ancestral pigment into the four groups. As 15. Felsenstein, J. (1985) Evolution 39, 783-791.
already described, lamprey rhodopsin shows an intermediate 16. Karnik, S. S. & Khorana, H. G. (1990) J. Biol. Chem. 29, 17520-17524.
profile in molecular net charge (Fig. 4). This is consistent with 17. Ovchinnikov, Y. A., Abdulaev, N. G. & Bogachuk, A. S. (1988) FEBS
the topology of the phylogenetic tree and suggests that a Lett. 230, 1-5.
molecular net charge has varied from basic to acidic in the 18. White, A., Handler, P. & Smith, M. L. (1964) Principles ofBiochemistry
course of evolution from cone to rod pigments. (McGraw-Hill, New York), 3rd Ed.
19. Yokoyama, R. & Yokoyama, S. (1990) Vision Res. 30,807-816.
It is reasonable to consider that the ancestral visual pig- 20. Yokoyama, R. & Yokoyama, S. (1990) Proc. Natl. Acad. Sci. USA 87,
ment is similar in character to cone visual pigments rather 9315-9318.
than vertebrate rhodopsins. Rhodopsin of the lowest verte- 21. Hisatomi, O., Iwasa, T., Tokunaga, F. & Yasui, A. (1991) Biochem.
brate (lamprey) and that of higher vertebrates diverged much Biophys. Res. Commun. 174, 1125-1132.
later than the divergence of the cone pigments into the four 22. Ovchinnikov, Y. A., Abdulaev, N. G., Feigina, M. Y., Artamonov,
groups. Taken together, we propose the hypothesis that I. D., Bogachuk, A. S., Eganyan, E. R. & Kostetskii, P. V. (1983)
animals had acquired the ability to distinguish color at least Bioorg. Khim. 9, 1331-1340.
at the stage of the lowest vertebrate and acquired scotopic 23. Hargrave, P. A., McDowell, J. H., Curtis, D. R., Wang, J. K., Juszczak,
vision later. E., Fong, S.-L., Mohanna Rao, J. K. & Argos, P. (1983) Biophys. Struct.
Mech. 9, 235-244.
We are grateful to Mr. Y. Imamoto and Mr. H. Imai for assistance 24. Nathans, J. & Hogness, D. S. (1984) Proc. Natl. Acad. Sci. USA 81,
in preparing chicken green, to Prof. T. Miyata and Mr. N. Iwabe 4851-4855.
(Department of Biophysics) for helpful discussions on the construc- 25. Sober, H. A., ed. (1970) Handbook of Biochemistry: Selected Data for
tion of a phylogenetic tree, and to Prof. T. G. Ebrey (University of Molecular Biochemistry (Chemical Rubber, Cleveland, OH), 2nd Ed., p.
Illinois) for a critical reading of the manuscript. This work was B-74.
supported in part by Grants-in-Aid for Scientific Research from the
Japanese Ministry of Education, Science and Culture, by a special 26. Konig, B., Arendt, A., McDowell, J. H., Kahlert, M., Hargrave, P. A.
& Hofmann, K. P. (1989) Proc. Natl. Acad. Sci. USA 86, 6878-6882.

27. Fukada, Y., Okano, T., Artamonov, I. D. & Yoshizawa, T. (1989) FEBS
Lett. 246, 69-72.

28. Al-Ubaidi, M. R., Pittler, S. J., Champagne, M. S., Triantafylios, J. T.,
McGinnis, J. F. & Baehr, W. (1990) J. Biol. Chem. 265, 20563-20569.

29. Sz6l, A., Takacs, L., Monostori, 9., Diamanstein, T., Vigh-Teichmann,

I. & R6hlich, P. (1986) Exp. Eye Res. 43, 871-883.
30. Bowmaker, J. K. & Knowles, A. (1977) Vision Res. 17, 755-764.
31. Oishi, T., Kawata, A., Hayashi, T., Fukada, Y., Shichida, Y. &

Yoshizawa, T. (1990) Cell Tissue Res. 261, 397-401.
32. Goldsmith, T. H. (1990) Q. Rev. Biol. 65, 281-322.